Deep_Learning/tensorflow_tSNE.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loss (hidden nodes : 5, iteration: 2000): 5.75\n",
      "loss (hidden nodes : 10, iteration: 2000): 32.62\n",
      "loss (hidden nodes : 20, iteration: 2000): 5.61\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f528fcfaf90>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAENCAYAAAD+CUlOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3Xl8FOX9wPHPd3Y3d7jDfQuCIohAxVu8FavUG8WK1rvW\nq2rr0V+1rVTbarVWrcV64C0eRa2KioqCiggIhEPkxiCBcAVCzt19fn88k2QDSXY3ye4m2e+b1752\ndnZm9ptJmO8+xzyPGGNQSiml6uMkOgCllFLNnyYLpZRSYWmyUEopFZYmC6WUUmFpslBKKRWWJgul\nlFJhabJQSikVVlyShYj0EpFPRWSZiCwVkRvd9feIyEYRWeg+xobsc4eIrBKRFSJySjziVEopVTuJ\nx015ItIN6GaMWSAi2cB84GfA+UCRMeaBvbY/EHgZOBToDswA9jfGBGIerFJKqX144/EhxphNwCZ3\nebeILAd61LPLOOAVY0wZsFZEVmETx1d17dCpUyfTt2/fpgtaKaWSwPz587caY3LCbReXZBFKRPoC\nhwBfA0cC14vIJcA84BZjzA5sIpkTslse9ScX+vbty7x582IRslJKtVoisj6S7eLawC0iWcAbwE3G\nmF3Av4D+wHBsyePBKI93lYjME5F5BQUFTR6vUkopK27JQkR82ETxojHmTQBjzGZjTMAYEwSexFY1\nAWwEeoXs3tNdV4MxZrIxZpQxZlROTthSlFJKqQaKV28oAZ4Clhtj/h6yvlvIZmcBS9zlt4HxIpIq\nIv2AgcDceMSqlFJqX/FqszgS+DmQKyIL3XV3AheKyHDAAOuAqwGMMUtFZCqwDPAD12lPKKWUSpx4\n9YaaDUgtb71Xzz6TgEkxC0op1WIEg0Hy8vLYs2dPokNpkTIzM+nZsyeO0/DKpLj3hlJKqWht3boV\nEWHQoEGNuuAlo2AwyMaNG9m6dSudO3du8HH0rCulmr2dO3fSpUsXTRQN4DgOXbp0obCwsFHHSfqS\nxeY9m3nt+9cAEBEq/1VWmlWtcd+r3K7qvdDlvV7XOKZU18KFHm+fY4Z+zl6fu/fxvOIlxZNCiieF\nVE8qqZ5UfB4faZ40OqV3ok1Kmxqfq1RLFQgE8Pl8iQ6jxfL5fPj9/kYdI+mTxZbiLUxePBlD65uL\nPN2bzrCcYUwYPIHjeh+X6HCUahT94tNwTXHukj5ZDM0ZyuKJi6teG2OqEkflssFQmUuq1pjq56p9\nQ9aHvq48Vm377/25odvtE0/ofgb8xk95oJyyQBllgTIqAhWUBcoo8ZewtWQrG4s2MmvjLG749AYe\nGvMQJ/Y5MRanUCmVBJI+WeytRrVPK/gic2vgVs5951yeW/acJgulWphLL72Ul156iZSUlKp1hYWF\neDyeuMeirUWtnM/j46geR7Fs2zIqghWJDkcpFaXf/OY3FBUVVT0SkShAk0VSGNBuAGWBMrYUb0l0\nKEq1On379uWBBx5g2LBhtG3blgsuuIDS0lIAZs6cSc+ePXnwwQfp3Lkz3bp145lnnklwxA2jySIJ\ndEzvCMC2km0JjkSp1mnq1KlMnz6dtWvXsnjxYp599tmq9/Lz8yksLGTjxo089dRTXHfddezYsQOA\nl156iWHDhtV77Mcff5wOHTowcuRI3njjjVj+GPXSNoskkOXLAqCooijBkSjVeH94ZynLftwV0884\nsHsb7j5jSMTb33DDDXTv3h2AM844g4ULF1a95/P5+P3vf4/X62Xs2LFkZWWxYsUKDjvsMC666CIu\nuuiieo/74IMP0rZtWz788EMuuOACunbtypFHHtnwH66BtGSRBNK96QCU+EsSHIlSrVPXrl2rljMy\nMigqqv5i1rFjR7xeb53v12fEiBFV+48dO5YJEybw5ptvNl3gUdCSRRLQZKFak2i+8bc2IlKju348\nackiCaR50wAo9ZcmOBKlVDRef/11ioqKCAaDfPjhh7zwwguceeaZCYlFk8WO9fD+b2HLd4mOJGa8\nji1ABoI6yrtSzcmLL77IkCF1l5T+8Y9/0KNHD9q1a8dtt93Gk08+yZgxY+IXYAithireBl8/Af2O\nhc6DEx1NTHjF/pr9pnFjwyil9rVu3boar++5556q5TFjxpCXl1fn9hMmTGDChAl1HnvWrFlNEWKT\n0JKFz9bn04rr8z2OvYlHSxZKqYbSZOHW51PReuvzPeImC51sUCnVQJosfBn2uRWXLKraLDRZKKUa\nSJOFL3lKFv6gtlkopRpGk4XjTqjSigfZc8T+mjVZKKUaSpOFW0VDK76QVs6qp9VQSqmG0mThcUsW\ngdabLMD2iNLeUEqphtJkIQLiadXVUGCrorRkoZRqKE0WYKuiWnE1FNhG7qAJJjoMpVQLpckCbFVU\nK6+GEhFNFkq1MFOnTuWII44gIyOj1mE+Fi5cyMiRI8nIyGDkyJE1hkZvaposwC1ZtO5qKC1ZKNXy\ndOjQgZtuuonbb799n/fKy8sZN24cF198MTt27GDixImMGzeO8vLymMSiyQLckkXrThaOOJoslIqB\nWE6reuKJJ3L++edXTawUaubMmfj9fm666SZSU1O54YYbMMbwySefNNnPFkqTBdgG7lbe+OuIQxBN\nFkrFQiynVa3L0qVLGTZsGCJSte7ggw9m6dKljfpZ6qKjzgI4Hgi27gupg5YsVCvx/u2Qnxvbz+g6\nFE67P+LNYzWtan2Kiopo27ZtjXVt2rRh9+7dDTpeOFqyALdk0bovpI7j6H0WSsVIrKZVrU9WVha7\ndtWci7ywsJDs7OxGH7s2WrIAe69Fa6+GwsGQmOkYlWpSUXzjb82GDBnCgw8+iDGmqipq8eLF/OpX\nv4rJ52nJAtxqqFaeLPSmPKVanEAgQGlpKX6/n2AwSGlpKRUVtjPOmDFj8Hg8PPLII5SVlfHII48g\nIhx//PExiUWTBSRFA7fH0a6zSjU34aZVff7550lPT+faa69l1qxZpKenc+WVVwKQkpLCtGnTeO65\n52jXrh3PPvss06ZNIyUlJSaxijGxr5oQkV7Ac0AXwACTjTH/EJEOwKtAX2AdcL4xZoe7zx3A5UAA\nuMEY80F9nzFq1Cgzb968hgX42GjIGQTnP9ew/VuAM/57BoM6DOKBYx9IdChKRW358uUccMABiQ6j\nRavrHIrIfGPMqHD7x6tk4QduMcYcCBwGXCciBwK3Ax8bYwYCH7uvcd8bDwwBTgUeF3EnZYgFcVp9\nNZTelKeUaoy4JAtjzCZjzAJ3eTewHOgBjAOmuJtNAX7mLo8DXjHGlBlj1gKrgENjFmAS9IbS4T6U\nUo0R9zYLEekLHAJ8DXQxxmxy38rHVlOBTSQ/hOyW566LDUdLFkopVZ+4JgsRyQLeAG4yxtToIGxs\n40lUDSgicpWIzBOReQUFBY0IrPWXLHS4D6VUY8QtWYiID5soXjTGvOmu3iwi3dz3uwFb3PUbgV4h\nu/d019VgjJlsjBlljBmVk5PTiOCcVt8bSpOFUqox4pIsxN4x8hSw3Bjz95C33gYmussTgbdC1o8X\nkVQR6QcMBObGLMAkuc9Ck4VSqqHidQf3kcDPgVwRqRw05U7gfmCqiFwOrAfOBzDGLBWRqcAybE+q\n64yJ4Vd/rYZSSql6xSVZGGNmA1LH2yfUsc8kYFLMggrlaLJQSqn66B3cYMeGSoZqKB2iXCnVQJos\nICmG+3BER51VqqW59dZbGThwINnZ2QwePJjnnqs5yoROqxpvSdLAraPOKtWyZGZm8s4771BYWMiU\nKVO48cYb+fLLLwGdVjUxkqFkgY46q1QsxHJa1T/84Q8MHjwYx3EYPXo0Rx99NF999VXVsXVa1XgT\np/U3cDsO8Rg0UqlkFI9pVUtKSvjmm2+qRqnVaVXjbOPOEnZv2UM/p4LURAcTQ1qyUK3FX+b+he+2\nfxfTzxjcYTC/PfS3EW8fj2lVr7nmGg4++GBOOeUUQKdVjbutu8tYs62UCr8/0aHElEc8WrJQKkZi\nPa3qbbfdxpIlS5g6dWpVSUKnVY0zjyMEcZBW/q1bRLRkoVqFaL7xtwZ3330377//Pp999hlt2rSp\nWq/TqsaZI0IQafVtFjrqrFItz3333cdLL73EjBkz6NixY433dFrVOPM4QoDW38Ct81ko1fyEm1b1\nzjvvZMOGDQwYMICsrCyysrL485//DMR/WlWthnJIimqoFE8KawrX8LvZv6Nndk8GtR/EET2OINXT\nmpv1lYq9devW1Xh9zz33VC2PGTOGvLy8OrefMGECEyZMqPPY4doZDznkEObPnx9xrI2R9MnCESFg\nHAi27m/dZw04ix92/cCXP35JQYmd+6NbZjf+c/J/6N2md4KjU0o1d1oN5VZDSSsfN2l0t9G8ePqL\nfHL+J3wz4RseO+Ex9lTs4b659yU6NKVUC5D0ycIRoRwvTqA00aHETZo3jWN6HsO5+5/LVz9+Rak/\neX52pVTDJH2y8DjCHtLx+Yshye5D2L/9/gRMgI1F+0xCqJRSNWiycIQik45j/OAvS3Q4cdUr285c\nm7c7L8yWSqlkl/TJwhGhiDT7oiw2t8k3V+1T2wNQWF6Y4EiUCk9HIGi4pjh3SZ8sKksWAJQnV7LI\nTMkEoKg8uqEHlIq3tLQ0tm3bpgmjAYwxbNu2jbS0tEYdJ+m7znpEKMJNFklWssj02WRR7C9OcCRK\n1a9nz57k5eVRUFCQ6FBapLS0NHr27NmoYyR9shCHkGSRXN+wUxx7p2d5IDaTpSjVVHw+H/369Ut0\nGElNq6EkpBoqyUoWIoJXvFQEKxIdilKqmdNk4Qh7Khu4k7Du3ufx4Q+27uHZlVKNl/TJwhFht8mw\nL8p21b9xK6QlC6VUJJI+WdQoWSRZmwXYkkVFQJOFUqp+SZ8sHIHiyglVk6zNAsDrePEbrYZSStUv\n6ZOFiCDi4BcfJGGvIJ+jJQulVHhJnyzAnVpVvJCEDb2OODrdqlIqLE0WuHNaJGnJQqdbVUpFQpMF\n7pwW4oEkrI5xxNFkoZQKS5MF9sY8Ww2lyUIppWqjyQJwHHEbuJMzWWibhVIqHE0WJHc1lLZZKKUi\nocmCkAbuJK2G0pKFUiqcuCQLEXlaRLaIyJKQdfeIyEYRWeg+xoa8d4eIrBKRFSJySqzj8zgQIHlL\nFjpHgFIqnHiVLJ4FTq1l/UPGmOHu4z0AETkQGA8Mcfd5XEQ8sQzOI0JAvEmZLERESxZKqbDikiyM\nMZ8D2yPcfBzwijGmzBizFlgFHBqz4HAbuPHqfRZKKVWHRLdZXC8ii91qqvbuuh7ADyHb5LnrYsbj\nCH69g1sppeqUyGTxL6A/MBzYBDwY7QFE5CoRmSci8xoz3aJHJKnbLLRkoZQKJ2HJwhiz2RgTMMYE\ngSeprmraCPQK2bSnu662Y0w2xowyxozKyclpcCxOZckiCauh9KY8pVQkEpYsRKRbyMuzgMqeUm8D\n40UkVUT6AQOBubGMxSNum0WSVkNpslBKheONx4eIyMvAGKCTiOQBdwNjRGQ4YIB1wNUAxpilIjIV\nWAb4geuMiW2lughUkJy9obTNQikVibgkC2PMhbWsfqqe7ScBk2IXUU0eRwhobyillKpTontDNQse\nR6jAk7TVUFqyUEqFo8kCO9xHRbKWLBy9g1spFZ4mC9z7LJK066ygd3ArpcLTZEFlb6jkrIbSNgul\nVCQ0WQCOAxUk57SqjuMQCGrJQilVP00WhDRwJ2E1lJYslFKR0GSB28BtPICBJPuW7YhDEE0WSqn6\nabIgpGQBSVcV5YhDMKjJQilVP00W2AZuW7Ig6aqi9D4LpVQkIk4WIvJrd3gOROQwEdkgImtF5PDY\nhRcfjiOUV97MnmTJQtsslFKRiKZkcTOw1l2+D/g7cC/wcFMHFW81SxbJVw21o2wH6wrXUVxRnOhw\nlFLNVDRjQ7U1xhSKSDZwMHCiMSYgIlHPQ9HceByhmFT74u+DIaMTpLe3j4wO0GE/6DkKBp0GvvTE\nBtvEumR0AeCMaWcA0DWzK1ccdAUXDL4gkWEppZqZaJLFDyJyBHZu7M/dRNEGaPEV3o4jzPQcwR1j\ne8LuTVC8HUp2QMl2KMyDNTNhzmOQmQPnPwd9jkh0yE3msoMu4+Ccg9m0ZxObizcze+Ns7v36XjyO\nh3P3PzfR4SmlmoloksVtwOtAOXCOu+6nxHiuiXjwCJSYFDj0yto3CPhh/Rfw7q/h1Yvhhm8hrW18\ng4wRRxxGdR1V9fqyIZdx3SfXMWnOJFI9qZyx3xkJjE4p1VxE3GZhjHnPGNPdGNPXGDPfXf0acGZs\nQosfxxECwXoG0/N4of+xcNZkKN4GS6fFL7g48zgeHjjmAUZ0GcGds+9k8uLJOtCgUiqq3lAHikgX\ndzlLRP4A3An4YhVcvHhECEZyQewxwlZFbfgq9kElUFZKFk+c+ASn9z+df377T/7w1R/wJ+G4WUqp\natH0hnoZaOcuPwAcAxwG/Lupg4o3T7iSRSUR6HwgbFsd+6ASzOfxcd9R93HF0Ct4Y+Ub3PbZbZQn\nWU8xpVS1aJJFX2PMChER4GzgPOBc4JSYRBZHjhNhyQKgTXfbCJ4ERIQbR9zIb3/yW2ZsmMENn9xA\nib8k0WEppRIgmmRR6nabPRTYYIzZCpQBaTGJLI4cIbKSBUB2N5sskmiIjIsPvJg/HvFHvvzxS675\n6BqKyosSHZJSKs6iSRYvAZ8AU4Bn3XUjqL5Rr8WybRYRbpze3s57kWQXzLMGnsVfj/kriwsWc8WH\nV7CzdGejjzl/83y2lmxtguiUUrEWTW+om4G7gGuNMY+6q4PYO7tbNMcRgpFmi9Qs+5xkyQLg1H6n\n8o/j/8HKHSu57IPL2Fi0sVHHu3T6pRw39TjtbaVUCxDVQILGmA+B1SJyuIj0NsbMM8Z8EqPY4sYj\nQiDSC1ZKtn0uS75kAXBMz2P414n/In9PPme/dTavfvdqgyZPCh2P6tLpl+r4VEo1c9F0ne0mIp8B\nK4E3gVUi8pmIdI9ZdHEScW8oAJ/bROMvjV1Azdyh3Q7lzTPfZFjOMO79+l7OfedcZuXNiqqEUBpy\n/hZsWcDBzx3Mt1u+jUW4SqkmEE3J4l/AIqCDMaYb0B5YCDwRi8DiKareUN7KZFEWu4BagG5Z3Zh8\n0mQePPZBygJl/PLjX3Lhuxcyfd30iO7JKAvse/4uef8Shk4Zypcbv4xFyEqpRohmuI+jgG7GmAoA\nY8weEfkN0LiK62bAI1GULDwp9rmWi12yERFO7nsyx/U6jmmrpzFl6RRu++w2emT14Lz9z2PcgHF0\nSu9U677F/rpHuL16xtUAHNTxIB45/hFyMnJiEr9SKnLRlCx2AAfutW4Q0PhuMQlmSxZEVo3idUen\nTfKSRSifx8d5+5/HW+Pe4uExD9M1sysPL3iYk147iV/P/DVf/vjlPm0SO8vC/9ks2baE4187nqFT\nhjJ0ylCWbl0aqx9BKRVGNCWLvwIzROQpYD3QB7gM+L9YBBZPHhEAgsYOKlgvTRZ18jgeTuhzAif0\nOYE1hWt48/s3eWv1W3y0/iN6ZPXgjP3O4Iz+Z9C7TW92lO6I+vjj3x1ftfzK6a8wpNOQpgxfKVWP\niJOFMeZJEVkNXAQMA350l4+NUWxx43HLV4GgweOEyRZV1VA69EV9+rftz60/uZUbRtzAjPUzmLZq\nGv9e9G+eWPQEh3Q+BCFcVq5faOKYef5MOqZ3bGzISql6RFOywO0mW9VVVkRSgQ+B3zdxXHHlOJUl\niwiqoRz3lOnAehFJ8aQwtv9YxvYfS/6efN5d8y5vr36bNYVrADiy+5Ec2+tYCooLmLd5XoN6RI2Z\nOgawc3PcPOJmRBqXiJRS+4oqWdShxf/PrKyGiqiRuypZtPg5n+Kua2ZXLh96Ob846Bcs27aMimAF\nwzsPr3XbPRV7mDRnEu+seSfi4z+z5BmeWfIMAJ+e/2mdjetKqehFdVNeHVr87beVVU8R3ZhXlSwq\nYhhR6yYiDOk0pM5EAZDpy+TPR/+Z3Im5zJ0wl5FdRkb1GcdNPY6hU4by8YaPGxuuUooIShYicnw9\nb6c0YSwJ41Q2cEdVstBqqHhJ96bz7KnPArB5z2ZOfP3EiPe96dObALhg0AXcNfouraJSqoEiqYZ6\nKsz7G5oikESqKllEkiw87lxPmiwSoktmF3In5gIw84eZXP/J9RHt9+qKV3l1xavs13Y/XjvzNXxO\ni5+zS6m4CpssjDH94hFIIjkNqYYKaLJItDG9xpA7MZeKYAXXf3I9X2z8Iuw+qwtXM+L5EQB8M+Eb\n0rwtfoR9peKiKdoswhKRp0Vki4gsCVnXQUQ+EpGV7nP7kPfuEJFVIrJCRGI+uVLVfRaRjGWn1VDN\njs/x8cSJT5A7MZc3z3wz4v1+8uJPGDplKIVlhTGMTqnWIS7JAjv/xal7rbsd+NgYMxD42H2NiBwI\njAeGuPs8LiKeWAZXdZ+Fdp1t8Qa2H0juxFwWX7KYiwZfFNE+R71yFEOnDNW5NZSqR1yShTHmc2D7\nXqvHYSdSwn3+Wcj6V4wxZcaYtcAq7Ox8MdOwBm7tDdWciQh3jL6D3Im5fHDOBxHtU9mDakvxlhhH\np1TLE6+SRW26GGMqJ7POB7q4yz2AH0K2y3PX7UNErhKReSIyr6CgoMGBRNXArfdZtDjds7qTOzGX\nRZcs4qYRN4Xd/oTXTmDolKHk78mPQ3RKtQyJTBZVjB3BL+r7NYwxk40xo4wxo3JyGj4yaXT3WTgg\njlZDtUCOOFw+9HJyJ+Yye/zssNuf9PpJDJ0ylN3lu+MQnVLNWyKTxWYR6QZ2YiWgsuy/EegVsl1P\nYjwMukRTDQW2dBHQaqiWrG1qW3In5pI7MZeHxzxc77ZHvHwELy5/MU6RKdU8JTJZvA1MdJcnAm+F\nrB8vIqki0g8YCMyNZSCho85GxPFpyaIVOaHPCeROzGXBxQsY1mlYrdvcP/d+FmxeEOfIlGo+4tV1\n9mXgK2CQiOSJyOXA/cBJIrISONF9jTFmKTAVWAZMB64zxsS0gSB01NmIOF5NFq2Qz+PjxdNfJHdi\nLlNOnbLP+xOnT6xlL6WSQ1MMJBiWMebCOt46oY7tJwGTYhdRTVW9oSKdWtXxaLJo5UZ0GUHuxFy2\nlWyrGtVWqWTWLBq4E83nFi0qApHclYcd8kOTRVLomN6RxZcsrnpd4i9JYDRKJY4mCyDVa09DmT/C\nZOF4dbiPJCIiVZM1bS/d+3YhpZKDJgsg1WdvEC+tiLBpRKuhks59R98HQKm/NMGRKJUYmiyANJ89\nDaUVkZYstBoq2WT5sgAorihOcCRKJYYmCyDNLVmU+SMtWXh1uI8kk+HLALTNQiUvTRZUJ4uIq6E8\nPh3uI8mke9MBTRYqeWmyANK80VZDaZtFssnw2pJFsV+roVRy0mRBA0oWOtxH0tGShUp2miwITRba\nwK1q53Hs34hff+8qSWmywI462ybNy9aissh2cLzaZpFkPO78W0ET4RcKpVqZuAz30RL06pDBhu0R\n1kc7HsibB0unQWYnaNsL2vUGd9gQ1fpUJotAbIcpU6rZ0mTh2i8niy9Xb8UfCOL1hClw9TsG1n4G\nr4UMLNehP5w8CQaPjW2gKiEcx/5NBLREqZKUVkO5xg7tytaicmavimAe5mNuhVu+h2u+gJ9Pg7EP\ngC8DXrkIlr0Vfn/V4njFfq/SaiiVrDRZuI4b3JnsNC9vL/oxsh2yu0DXg2C/4+DQK+GKGdBzFEy7\nDgpjOleTSgBH7H8Vv9EGbpWcNFm4Ur0eThnSlY+WbaY80gEFQ/nS4Zz/QKAcPvlT0weoEkobuFWy\n02QR4tQhXdld6ufL1RFURdWmfV9bylj8Kuz8oUljU4lV2XVWG7hVstJkEeKogZ3ITo2iKqo2o6+2\nz/OebpqgVLNQWQ2lDdwqWWmyCJHm83D6sG5MX5LPnrIG1k236w37nwbfvqBzXrQyHvFoNZRKWpos\n9nLuyJ4Ulwd4L3dTww9y8HjYswXWfd50gamE84hHq6FU0tJksZeRfdrTt2MGbyzIa/hBBp4MqW0g\n942mC0wlnMfxaDWUSlqaLPYiIpwzoidz1mznh0jv6N6bLw0G/xSWvwP+CIcQUc2eI46WLFTS0mRR\ni7NH9kSExpUuhp4DZYWw8sOmC0wllFZDqWSmyaIWPdqlc3j/jryxII9g0DTsIP3GQGYOLJ7apLGp\nxNEGbpXMNFnU4dyRPflhewnfrNvesAN4vHDQOfD9B1Ba2LTBqYTQaiiVzDRZ1OHUg7qSmeJpZFXU\n+RAog2VvN11gKmG0gVslM00WdchI8TJ2aDfeXbyJ4vIG3i/RYwR02M/e0a1aPG2zUMlMk0U9zh3Z\nkz3lAaYvyW/YAURg2PmwbjYUNqKEopoFRxxts1BJS5NFPX7StwN9Ombwwpz1DT/IwRfapKHDf7R4\nXser1VAqaWmyqIfjCJcd0ZcFG3Yyf/2Ohh2kfR8YNBbmPQMVJU0boIorbeBWyUyTRRjnjepFdpqX\np2evbfhBRl8NJdsh97WmC0zFnXadVclMk0UYmaleJozuw/tLNrEif3fDDtL3aOg6DGY9CP7ypg1Q\nxY1HPDr5kUpamiwicPUx/clM8fK3D75r2AFE4Pj/gx3r4NvnmzQ2FT/awK2SmSaLCLTPTOGaMfsx\nY/mWhk+MNPAk6H04fDoJ9mxr2gBVXGjXWZXMEp4sRGSdiOSKyEIRmeeu6yAiH4nISve5faLj/MWR\n/ejdIYM73sylpLwBFwwROP3vULoLpt/e9AGqmNOb8lQyS3iycB1njBlujBnlvr4d+NgYMxD42H2d\nUOkpHu4/ZyjrtxXzl+kNrI7qciAcfQvkToVFrzRtgCrmtIFbJbPmkiz2Ng6Y4i5PAX6WwFiqHLFf\nJy49oi/PfrmOtxZubNhBjrnVNni/fQPkzWvaAFVMaddZlcyaQ7IwwAwRmS8iV7nruhhjKqeqywe6\n1LajiFwlIvNEZF5BQUE8YuWu0w/g0H4d+M3ri5nXkEEGPT44bwq06QbPnw1585s+SBUTWg2lkllz\nSBZHGWOGA6cB14nIMaFvGmMMNqHswxgz2RgzyhgzKicnJw6hgs/j8PiEEfRol86lz3zDgg0NuFkv\nsyNM/B/AiNfFAAAbqElEQVRktIfnzrSTJKlmT6uhVDJLeLIwxmx0n7cA/wUOBTaLSDcA93lL4iLc\nV6esVF668jA6ZqVw0ZNzeL8h83W36wWXvQ85g+DVi+HdW23jt2q2tBpKJbOEJgsRyRSR7Mpl4GRg\nCfA2MNHdbCLwVmIirFvXtmm8fs0RHNitDde+uIB7/7eM0oooLyRtutuEMfpa+OY/8NihMH+K3rjX\nTHnFq8kiFoyBZ06HFe8nOhL7f0///9Uq0SWLLsBsEVkEzAXeNcZMB+4HThKRlcCJ7utmJyfbljAu\nPqw3/5m9lrGPzOLT77Zga84i5E2F0+6HKz+2yeOdG+CfI+DzB2DXj7ELXkVNSxYx4i+F9bPh1Z83\n7XGNgTn/gqK9KiYCFfDeb2D35up1hXnw5T/hr/3hgYHhj7t0GvjLGhZXaDJaOwu21zGUUFkRfPyn\nZpO8EposjDFrjDEHu48hxphJ7vptxpgTjDEDjTEnGmMaOF1d7KX5PNz7s6G8cPloAkHDZc9+w/jJ\nc/j0uy3RTcnaYyRc8TFc/Aa07wuf/AkeGgJTzoSvHodtq2P2M6jIeBwPwWATtFkYA8vegvoayytK\n7IUkWhWl9mJYn3WzYetKWDXDXowAClbAty/U3C5QAZH8vMvetvs3VGW8wZC4Ny+Fogg6rXz5T/ju\nXfvFqnJGyuXvwDdPQcF39p6m139Rc59VH8Pcf8N7t1ave/lC+PB3UL4bSndWrw8GYPJx8P2H1evW\nzITXJsIbV0T/hW79l3Bvjv0dAEz5KTwyvOY2sx+C796DWQ/Yx5Sfwo8L3Z/3Ufj0z9F9ZhNJdMmi\n1ThqYCc+uvlY/jhuCGu37uGyZ7/h+Adn8tinq1i3dU9kBxGBASfCpf+DG76192TszocP7rCljYeG\nwuuXw9eT4cdvG/7NJhltnL9vm9A7N9mLTW1mPwyLaw786BUvpUX5UPC9XfHlo9WdE/xlUB7m91y6\nC8qL7bzsUy+BuZOr31v2ds39373FXiQKVtiLXeVnAqz93I4CsGdr9bfjWQ/aC8qkLvDk8XZdeTG8\ncWXNb9AAz54Oj46CF86xFyOAJ46Ct66rud2fOtmS7rov7GcBfPUY3NO2ZhKZ+nNbhVrJGPj2RfuN\n+JN7YeZf7Pqy3fDnHvZivWO9PU5+LgRrGW/rX0fA46NrP48r3q9OTh/+Dl65CP5+ADx2mP3MVy+G\nd39dnYzXzYLi7fb9f46qPo6/DOY8Yc97WS3thRUl9uf+cQG8eSXMfdImocpksvxt+7nrv4L3a7kV\nzBj44h/2GMGgfVQmidWf1v6zAcy4B165sPr/9w9fw+Rj3Z/3LvjsL9Xbvng+vDKh7mM1IW9cPiVJ\npHgdLjm8L+N/0pv3l2zihTnr+dsHK/jbBysY0r0NJwzuzNH75zC8Vzt8njB5ukN/OP539rFjnf1m\ns/4L+81kyet2G/FAxwH2Zr/OB0LOYLtfh36Qkhnzn3cfRQXgTYG0tvH/7Lrs3gwZHe0FtO/RcPZk\nexFp1wvmP2O3GXU5OF57cRpwEpx8L8y427437LyqQ/UOCu8H9lDxwtn4bl5i/+MC3FMI/z4WCpbD\nDQvh9cvg2N/CoNPsBWPtZ9DvWLi/F2R1hSOut/tVToi1abG94B58IZz1hE1s+bn2vdWfwJI37LaX\nf2hjn3KGHZgyf7Hd5u6d8PEf7QOq1/+5m332psC4x+zyj9/Wfp4CblVHoMJ2766sSv32efvofCD8\n9CH44E53uzJw0mseo3i7/bv77l1465ewcz18/jf73pjfwuZlUF4EM++DIWe7x38Bjrq59piKt9lj\nfjrJbuNNs1+eXh5ffd5D7f7RfmuvJCH/x14eb38/YL+UAaz8wD62r665LcDCl2DatdC+n31tgtUl\nkfOfq7ntM6fa51Pvs8feMMd2XNm6Cj76vX10HWp//uPc82cC8MPc2n/u2uLfW6DClmpWflD/MZqQ\nJosYSPE6jBveg3HDe7BxZwnv527ivdxNPPrpKh75ZBVZqV4O368jRw/sxNEDc+jbMQOp/AOuTfu+\nMPoq+wB74cj7xhbVNy+z3yiXTqNGD+PMzjZptO9nn9v1hjY93Ed3SMlo+h/8gQH2+cbFdh6PRHvv\nNvvtPd0dLWbdLPtNEKovVlB9UQVbPTH33/seK/d1Oix4ATp1YFegjI5L/1v93pqZ1ReiyiqFl8fb\n5NRlCHz9RPW2RfmwwL3fdMMc++260s4N9kL7ykXV67ZWlijcv48S91vt5iXV28z5177xfvloyAt3\n34//aEsge1v4UvVyRYlNnO/dVnObLcvst/hKgQrwpdsSUKW/9oPUNtXf0kO/AS//X3U1EULV3+qG\nOZAf8rOs+az6Yg7w0f/ZhPLNf+zf9J6Q9odFYaYrXv9FSPzLq5dXf1Jzu3lP71u6mXatfd7htieE\ndpnevLT2zyvZYUsBL4+HzkPs/9tKlcl/+xr3uOvhqZNq7h8M2tEdKhXsNVLEd+9VL7/7a1iwV9KK\nMYmqMbYZGzVqlJk3r3nfEV1YXMGXq7fy+cqtzFpZQN4OOxlSj3bpHLN/J44akMORAzrSLiMl+oOX\n77EXlu1r7R/49rXVy7tquds8vYNNHG3d5FGZRLI6Q1YX+8joCI4n8hjuqaNEkd7e/mcr3UUdt8w0\nX7esgEcPhbJCPsxI55YuOTz742ZGlsWgCjAzB/bUUU/f+3CbfD7/a/THHXEJnHIf3Ncj/LZnP2mr\nQv/ar/7tLn0PNi2sLmmoxrnmC1vVtWVZw/bfu5QVBRGZHzLUUt3babJIDGMM67cVM2tlAbNWbuWr\n1dvYXea303b3aMuYQZ356bBuDOyS3fgPqyi1CWPXRlt0LcwLWXbXl9TSh0AcyOjkJo8c97mz/YaX\n1QUyOkBaO0hrY6su/nFw42NtxjZ4vZzeqzu/37qN83ZH2A6lVDzcuKhmSSYKmixaGH8gyKK8ncxa\nuZVZK7eyYMMOjIFBXbL56bBunDOyJ93bpYc/UEOVF8PuTfabbdFm292waEv18p6Q14Hm0ZUv3kpE\nOLRvL87cXcSkrc22g55KRtnd4Zbl4berhSaLFm7L7lLez83nf4t/5Jt1O3AETjigCz8/rA9HDeiE\n49TTxhGBmSu2cOkz3/D2r45kWM92ke9ojK17LtpsGx/LdtnqpbJC25jcpjv0Gg2eFFuHW7Ldbr9l\nme3+u2Ntdf1tC3Rk7x7s8njIXbsh0aEoVVMDq6IiTRbawN1Mdc5OY+IRfZl4RF9+2F7MS3M38Oo3\nP/DRss0M6pLNdccP4PSh3fA0MGn8Zbrtenjmo1/w1MRRnHBArWM17ksE0tvZRywZY7sOlu2ySaei\n2Cajsl32UZhn22mKttgSUVmRfa6tOq2S46vZl78BUt0vV4aqZmOlkoKWLFqQMn+Adxdv4vGZq1m1\npYj+OZnceMJAzhjWPeqSxvEPzmRNQXW9+zu/OoqhPZtRl9emVFYEGEh123+Msd1LS3ZA32NsF81d\nG+3d9P4y+156Bzjgp3b7battIvr2BW73FvHu9kU8PeASflJaCsfdBQtfgO8/gDF32BumyotsF9I9\nBbZUVbLTVuNldYExt8P/boaBJ9sb4yp724gHrphhu8WWF9l1OYNtj5iuQ2HC6/Z+i/Vf2O63Hfez\ny9ndbZfRUG16wq686tdDz7M3+BXlx/Q0qwTyZcJdDRvxQauhWrFg0DB9aT6PfLyS7/J3M7RHW+4c\newCH79cx4mMcef8nbNxZUmPdG9cezsg+HZo63FZlXv48LvvgMoZ2GspLp78UfodI+MttbzFfWvW6\nPVttwjJBwNgOBGC7V5qg7aW260fbm23VDFvK6n0EbM6F/Y633XnXfQGHXWs7IoC9f+PzB6HHIbbr\n8ILn7PwqlcNiDDnLdqlu29N209y0CA480ybTdn1ssvpxAVw/3yasNTPBX2JvSjvoHJt8Z//dxjVo\nrO3dtfRNOPJGe6NZZg70PcreIzJ3Moy+Gratst152/SEGxbY7eY8buMdcwfMfxa6DYfvQ8aNatcH\nDrnY3n+xtzY9bOI/7a/w/m/suoPOhTWf2p/j6Ftq7z4cqQPH2bvvK3nT7HAl/Y6xN0smylE3w4n3\nNGhXTRZJIBg0vLVoI3+bvoIfC0s5YXBn7hg7mAGdw/egGnXvDLYW7dv9c+6dJ9C5TVoteyiAikAF\nI14YAUDuxJbb9tKsbF1pu2lndLCJa8sy6DQIPGFqyfPm21JYens3QZ1q99+53vYMKvjelhYr7/kx\nxlajblwAvgx7z8P+p9hEu/xte0Pr7nybNAu+syW73fn2/Q79oWKPLZ0ue8v2CPSlQfdDbPVodhd7\nw+P2tTah7PoRlk2zN6gecKatNg2UwWd/g0OvhLa97N3d7fvYC32gwt7h/tlfoO+RsP9p9sbQ7G72\nPqn9TrD3uaRkwYl32zvpF70Mo35hvxA0giaLJFJaEeCZL9bx+KerKK4IMGF0b246cX86ZNZ9v8bQ\nuz9gd1ktwywAM28dQ99OCbgDvIUYOmUoAHMumkOmT8+TatkiTRY6NlQrkObzcO2Y/Zh52xguOrQ3\nL369gWP/9in/mbWGcn/tA8HVlSgAxjwwk763vxv9kOtJom+bvgA8tzS+d9AqlUiaLFqRjlmp/Oln\nBzH9xqMZ2ac99767nJMe+ozpS/KjGzbdNfj/pnPqw58TiGb03CTwt2PteEePL3o8wZEoFT+aLFqh\ngV2yefayQ5nyi0NJ8Thc88J8xk+ew5KN0ffD/i5/N/vd+R63vbaoQQmnNRrcYXDV8s7Q4ayVasU0\nWbRix+6fw/s3Hs29PzuIlVuKOOPR2dwydRGbd5VGfazX5ufR7473eOij78NvnESOfvXoRIegVFxo\nA3eS2FVawWOfruKZ2evwOEJJI9sjfnXcAG49ZVATRdfy5O/J56TX7aihX1/0NRm+GIziq1QcaG8o\nVasN24p58KMVfPLdFh6fMIKjB+bss01xuZ9z//UVyzbVMiHMXk4f2o1HLjykwXeSt2SVvaJAu9Gq\nlkuThWoSW4vKGHXvjIi2zb3nZLLTfDGOqPlYsX0F575zLgAZ3gy+nvB1giNSKnqaLFSTW/pjIac/\nMjvsdn8//2DOHtEzDhElXmjpIt2bztwJYWY/U6qZ0WShYurDpflc9fz8sNv97/qjOKhHKx1zCjsv\nybDnhtVY9+3Pv8Xr6BidqmXQZKHi5r/f5nHzq4vCbvfExSM4ZUjX+qeQbYEqghWMeH5EjXX/Ofk/\njO42OkERKRU5TRYqIWatLODnT4WviumcncrLVx3GfjlZcYgq9oImyMHP7TtT4Ozxs2mb2npLVqrl\n02ShEm5ncTlXPz+fr9dGNqvcn8YN4YKf9CbF23Jv/3lo/kM8veTpfda/d/Z79MrulYCIlKqfJgvV\n7PywvZhfvriA3CjuJP/Z8O5cd9wABnTOajHVV2WBMka9UPv/vcO7Hc4jxz9CmldH9lXNgyYL1ez5\nA0Gmzsvjzv827B6F647bj58N79FsE0mJv4RjXz2WEn9Jnds8f9rzDO88PI5RKVWTJgvVIhljWJxX\nyEMzvmfmioJGHeunw7px6kFdObhnO7q2TcPnSVz11pxNc7jywyvDbnfe/udx9bCr6ZIZ4TS3SjWS\nJgvV6hSV+fl4+Wae+WIdC39o+gH8jh7YicP368iwHu3o2ymDTlmppHqdJi+1bCnewi9n/JIVO1ZE\nvM8Z/c/grIFnMaTjEB1aRDUpTRYq6QSChi27S1m4YScfLM1n2sKGzUncGI7YoeIHds5iQOcs+nTM\npGf7dLq1TaNDZgpt0n2k+zx4HalKQhWBCt5Y+QaTvq5lmtAoHdnjSI7sfiTDcobRJ7sP2SnZeBxP\no4+rWi9NFkrVIxg07C7zs3FHCcs37eLrtduYtXIrmwqjH5E3dgI4afn42iwgpeMXTX50D17SPFmk\nezPJ8GaR6csg05dFpi+drJRMMn0ZZPgyyPRl0DY1mwM6HkCntI50SG9Puje9WbYTqehpslAqhowx\nBIKGkooAO4sr2FRYyobtxawpKGLVFvsoKCqjzB+sc7bCphFAvEU4aRvxZq7Em70Mxxf9vCVNyRgP\nuI/QZYxjX+OEvO8A9j2Mg6l6zwHEbuM+23W4z5X7OmAk5FlCjuO+xgk5ntTyXuVy5fu1rxc8CD6M\n8SDGi+B1Y/fiULnOwevxkeb1kOJ18HkcUjwOKV4Hr0eqln0eB68j+Lz2fa8jeD0OqV4HjyNV23oc\nqdrWW/UseB3B4zj4PEJmqjC6f1uyUxs2xW+kyULHJFCqAUTsf9psj0N2mo9eHTI4tF+HmH9uZZKq\nCBjK/AFKKgLsKfOzu9TPrlI/hSUV7CwuZ2dxBdv3lFe9LvMH8QcM/mCQioChIhAkEDT4gwEqgmX4\nzR78TiF+ZzsBZxuGMgL4MfgxBECCGH8WSBARP4gfJOAuB0KWgyGvA0DQ3TaISABx/EDl+0G7PUH3\ndev44mqAEvcRdsMK99FIneeO4eOJ/2z8geqhyUKpFqQySXk9kJ7ioV2iA0oAYwwVwQpK/CWU+cso\n9ZdR7C+mxF9CecBPRdBPIBhwE2GAQMA++0MeFcEA/qr1Qbs+ZLuKQICAu74iWL1stw1WHT/gvg4a\nQyAYIGCCBIJB+2yCBCvfJ0AwaOy6qoep8WxTZRDHt8MmXAyIAQwSslz1cJOrCaQx7uBjY37em3Wy\nEJFTgX8AHuA/xpj7ExySUirBRIQUTwopnhRITXQ0yaPZjqsgIh7gMeA04EDgQhE5MLFRKaVUcmq2\nyQI4FFhljFljjCkHXgHGJTgmpZRKSs05WfQAfgh5neeuU0opFWfNOVmEJSJXicg8EZlXUNC4oSGU\nUkrVrTkni41A6JjOPd11VYwxk40xo4wxo3JycuIanFJKJZPmnCy+AQaKSD8RSQHGA28nOCallEpK\nzbbrrDHGLyK/Aj7Adp192hizNMFhKaVUUmq2yQLAGPMe8F6i41BKqWTXasaGEpECYH0jDtEJ2NpE\n4TQljSs6Gld0NK7otMa4+hhjwjb6tppk0VgiMi+SwbTiTeOKjsYVHY0rOskcV3Nu4FZKKdVMaLJQ\nSikVliaLapMTHUAdNK7oaFzR0biik7RxaZuFUkqpsLRkoZRSKqykTxYicqqIrBCRVSJye5w/u5eI\nfCoiy0RkqYjc6K6/R0Q2ishC9zE2ZJ873FhXiMgpMYxtnYjkup8/z13XQUQ+EpGV7nP7eMYlIoNC\nzslCEdklIjcl4nyJyNMiskVEloSsi/r8iMhI9zyvEpFHpJETW9cR199E5DsRWSwi/xWRdu76viJS\nEnLenohzXFH/3uIU16shMa0TkYXu+nier7quDYn7GzPGJO0De2f4aqA/kAIsAg6M4+d3A0a4y9nA\n99i5O+4Bbq1l+wPdGFOBfm7snhjFtg7otNe6vwK3u8u3A3+Jd1x7/e7ygT6JOF/AMcAIYEljzg8w\nFzgMO9nz+8BpMYjrZMDrLv8lJK6+odvtdZx4xBX17y0ece31/oPA7xNwvuq6NiTsbyzZSxYJnTPD\nGLPJGLPAXd4NLKf+YdjHAa8YY8qMMWuBVdifIV7GAVPc5SnAzxIY1wnAamNMfTdixiwuY8znwPZa\nPi/i8yMi3YA2xpg5xv6vfi5knyaLyxjzoTHG776cgx2Us07xiqseCT1fldxv4OcDL9d3jBjFVde1\nIWF/Y8meLJrNnBki0hc4BPjaXXW9W23wdEhRM57xGmCGiMwXkavcdV2MMZvc5XygSwLiqjSemv+J\nE32+IPrz08Ndjld8AL/Afrus1M+tUvlMRI5218Uzrmh+b/E+X0cDm40xK0PWxf187XVtSNjfWLIn\ni2ZBRLKAN4CbjDG7gH9hq8aGA5uwReF4O8oYMxw7re11InJM6Jvut5SEdKUTOwrxmcBr7qrmcL5q\nSOT5qYuI3AX4gRfdVZuA3u7v+dfASyLSJo4hNbvf214upOYXkrifr1quDVXi/TeW7Mki7JwZsSYi\nPuwfw4vGmDcBjDGbjTEBY0wQeJLqqpO4xWuM2eg+bwH+68aw2S3WVha9t8Q7LtdpwAJjzGY3xoSf\nL1e052cjNauEYhafiFwK/BSY4F5kcKsstrnL87H13PvHK64G/N7ieb68wNnAqyHxxvV81XZtIIF/\nY8meLBI6Z4ZbJ/oUsNwY8/eQ9d1CNjsLqOyp8TYwXkRSRaQfMBDbeNXUcWWKSHblMraBdIn7+RPd\nzSYCb8UzrhA1vvEl+nyFiOr8uNUJu0TkMPdv4ZKQfZqMiJwK/AY40xhTHLI+R0Q87nJ/N641cYwr\nqt9bvOJynQh8Z4ypqsKJ5/mq69pAIv/GGtNi3xoewFhsT4PVwF1x/uyjsMXIxcBC9zEWeB7Idde/\nDXQL2ecuN9YVNLLHRT1x9cf2rFgELK08L0BH4GNgJTAD6BDPuNzPyQS2AW1D1sX9fGGT1SagAlsP\nfHlDzg8wCnuRXA08inujbBPHtQpbn135N/aEu+057u93IbAAOCPOcUX9e4tHXO76Z4Fr9to2nuer\nrmtDwv7G9A5upZRSYSV7NZRSSqkIaLJQSikVliYLpZRSYWmyUEopFZYmC6WUUmFpslAqwUSkyO23\nr1SzpclCJT2xw1CfKCKXisjsGH/WTBG5InSdMSbLGLMmlp+rVGNpslCqibhDRCjVKmmyUMo6AHgC\nONytFtoJ4A6f8ICIbBCRzSLyhIiku++NEZE8EfmtiOQDz4hIexH5n4gUiMgOd7mnu/0k7Eimj7qf\n8ai73ojIAHe5rYg85+6/XkR+JyKO+96lIjLbjWeHiKwVkdPifqZUUtJkoZS1HLgG+MqtFmrnrr8f\nO1jccGAAdnjn34fs1xXogJ2E6Srs/6ln3Ne9gRLsEAsYY+4CZgG/cj/jV7XE8U+gLXbIlWOxY/lc\nFvL+aOxwDp2wE+E85Y75o1RMabJQqg7uRfgq4GZjzHZjJ6H5M3bAyUpB4G5jRyQtMcZsM8a8YYwp\ndrefhL3oR/J5HvfYdxhjdhtj1mGH7f55yGbrjTFPGmMC2MlvulE9p4FSMaN1rErVLQfIAOaHfHkX\n7JSulQqMMaVVb4pkAA8BpwKVk/lki4jHvcDXpxPgA0Jn/1tPzclq8isXjDHFblxZkf5ASjWUliyU\nqrb3qJpbsdVIQ4wx7dxHW2NMVj373AIMAkYbY9pg53gGm2Rq237vz6vAVmFV6k2c51hRqjaaLJSq\nthno6c5tgqmelOchEekMICI9ROSUeo6RjU0wO0WkA3B3LZ9R6z0VbsljKjBJRLJFpA92RrYXGvEz\nKdUkNFkoVe0T7HwF+SKy1V33W+x8EHNEZBd2DoFB9RzjYSAdW0qYA0zf6/1/AOe6vZkeqWX/64E9\nwBpgNvAS8HTDfhylmo7OZ6GUUiosLVkopZQKS5OFUkqpsDRZKKWUCkuThVJKqbA0WSillApLk4VS\nSqmwNFkopZQKS5OFUkqpsDRZKKWUCuv/ASvsh/4TNqLpAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f528c843410>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import requests\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "IRIS_TRAIN = \"http://download.tensorflow.org/data/iris_training.csv\"\n",
    "IRIS_TEST = \"http://download.tensorflow.org/data/iris_test.csv\"\n",
    "\n",
    "names = ['sepal-length', 'sepal-width', 'petal-length', 'petal-width', 'species']\n",
    "train = pd.read_csv(IRIS_TRAIN, names=names, skiprows=1)\n",
    "test = pd.read_csv(IRIS_TEST, names=names, skiprows=1)\n",
    "\n",
    "Xtrain = train.drop('species', axis=1)\n",
    "Xtest= test.drop('species', axis=1)\n",
    "\n",
    "ytrain = pd.get_dummies(train.species)\n",
    "ytest = pd.get_dummies(test.species)\n",
    "\n",
    "def create_train_model(hidden_nodes, num_iters):\n",
    "\n",
    "    tf.reset_default_graph()\n",
    "\n",
    "    X = tf.placeholder(shape=(120, 4), dtype=tf.float64, name='X')\n",
    "    y = tf.placeholder(shape=(120, 3), dtype=tf.float64, name='y')\n",
    "\n",
    "    W1 = tf.Variable(np.random.rand(4, hidden_nodes), dtype=tf.float64)\n",
    "    W2 = tf.Variable(np.random.rand(hidden_nodes, 3), dtype=tf.float64)\n",
    "\n",
    "    A1 = tf.sigmoid(tf.matmul(X, W1))\n",
    "    y_est = tf.sigmoid(tf.matmul(A1, W2))\n",
    "\n",
    "    deltas = tf.square(y_est - y)\n",
    "    loss = tf.reduce_sum(deltas)\n",
    "\n",
    "    optimizor = tf.train.GradientDescentOptimizer(0.005)\n",
    "    train = optimizor.minimize(loss)\n",
    "\n",
    "    init = tf.global_variables_initializer()\n",
    "    sess = tf.Session()\n",
    "    sess.run(init)\n",
    "\n",
    "    for i in range(num_iters):\n",
    "        sess.run(train, feed_dict={X: Xtrain, y: ytrain})\n",
    "        loss_plot[hidden_nodes].append(\n",
    "                sess.run(loss, feed_dict={X: Xtrain.as_matrix(), y: ytrain.as_matrix()}))\n",
    "        weights1 = sess.run(W1)\n",
    "        weights2 = sess.run(W2)\n",
    "\n",
    "    print \"loss (hidden nodes : %d, iteration: %d): %.2f\" % (hidden_nodes, num_iters, loss_plot[hidden_nodes][-1])\n",
    "    tf.summary.FileWriter('ten', sess.graph)\n",
    "    return weights1, weights2\n",
    "\n",
    "num_hidden_nodes = [5, 10, 20]\n",
    "loss_plot = {5: [], 10: [], 20:[]}\n",
    "weights1 = {5: None, 10: None, 20: None}\n",
    "weights2 = {5: None, 10: None, 20: None}\n",
    "num_iters = 2000\n",
    "\n",
    "for hidden_nodes in num_hidden_nodes:\n",
    "    weights1[hidden_nodes], weights2[hidden_nodes] = create_train_model(hidden_nodes, num_iters)\n",
    "    plt.plot(range(num_iters), loss_plot[hidden_nodes], label=\"nn: %d\"%hidden_nodes)\n",
    "\n",
    "plt.xlabel('Iteration', fontsize=12)\n",
    "plt.ylabel('Loss', fontsize=12)\n",
    "plt.legend(fontsize=12)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy : 96.00\n",
      "accuracy : 93.00\n",
      "accuracy : 96.00\n"
     ]
    }
   ],
   "source": [
    "X = tf.placeholder(shape=(30, 4), dtype=tf.float64, name='X')\n",
    "y = tf.placeholder(shape=(30, 3), dtype=tf.float64, name='y')\n",
    "\n",
    "for hidden_nodes  in num_hidden_nodes:\n",
    "    W1 = tf.Variable(weights1[hidden_nodes])\n",
    "    W2 = tf.Variable(weights2[hidden_nodes])\n",
    "    A1 = tf.sigmoid(tf.matmul(X, W1))\n",
    "    y_est = tf.sigmoid(tf.matmul(A1, W2))\n",
    "    \n",
    "    init = tf.global_variables_initializer()\n",
    "    with tf.Session() as sess:\n",
    "        sess.run(init)\n",
    "        y_est_np = sess.run(y_est, feed_dict={X: Xtest, y: ytest})\n",
    "    \n",
    "    correct = [estimate.argmax(axis=0) == target.argmax(axis=0)\n",
    "              for estimate, target in zip(y_est_np, ytest.as_matrix())]\n",
    "    accuracy = 100 * sum(correct)/ len(correct)\n",
    "    print 'accuracy : %.2f' % accuracy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## t-SNE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from numpy import linalg\n",
    "from numpy.linalg import norm\n",
    "from scipy.spatial.distance import squareform, pdist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import sklearn\n",
    "from sklearn.manifold import TSNE\n",
    "from sklearn.datasets import load_digits\n",
    "from sklearn.preprocessing import scale\n",
    "from sklearn.metrics.pairwise import pairwise_distances\n",
    "from sklearn.manifold.t_sne import (_joint_probabilities,\n",
    "                                    _kl_divergence)\n",
    "from sklearn.utils.extmath import _ravel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "RS = 20150101"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.patheffects as PathEffects\n",
    "import matplotlib\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "sns.set_style('darkgrid')\n",
    "sns.set_palette('muted')\n",
    "sns.set_context(\"notebook\", font_scale=1.5,\n",
    "                rc={\"lines.linewidth\": 2.5})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1797, 64)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "digits = load_digits()\n",
    "digits.data.shape\n",
    "# print(digits['DESCR'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/root/anaconda2/lib/python2.7/site-packages/matplotlib/font_manager.py:1297: UserWarning: findfont: Font family [u'sans-serif'] not found. Falling back to DejaVu Sans\n",
      "  (prop.get_family(), self.defaultFamily[fontext]))\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAC3CAYAAACxII3nAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIiKJ/g+dmQECkMaaHQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f16634d3e50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "nrows, ncols = 2, 5\n",
    "plt.figure(figsize=(6, 3))\n",
    "plt.gray()\n",
    "\n",
    "for i in range(ncols * nrows):\n",
    "    ax = plt.subplot(nrows, ncols, i+1)\n",
    "    ax.matshow(digits.images[i, ...])\n",
    "    plt.xticks([]); plt.yticks([])\n",
    "    plt.title(digits.target[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = np.vstack([digits.data[digits.target==i]\n",
    "               for i in range(10)])\n",
    "y = np.hstack([digits.target[digits.target==i]\n",
    "               for i in range(10)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "digits_proj = TSNE(random_state=RS).fit_transform(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def scatter(x, colors):\n",
    "    # We choose a color palette with seaborn.\n",
    "    palette = np.array(sns.color_palette(\"hls\", 10))\n",
    "\n",
    "    # We create a scatter plot.\n",
    "    f = plt.figure(figsize=(8, 8))\n",
    "    ax = plt.subplot(aspect='equal')\n",
    "    sc = ax.scatter(x[:,0], x[:,1], lw=0, s=40,\n",
    "                    c=palette[colors])\n",
    "    plt.xlim(-25, 25)\n",
    "    plt.ylim(-25, 25)\n",
    "    ax.axis('off')\n",
    "    ax.axis('tight')\n",
    "\n",
    "#     We add the labels for each digit.\n",
    "    txts = []\n",
    "    for i in range(10):\n",
    "        # Position of each label.\n",
    "        xtext, ytext = np.median(x[colors == i, :], axis=0)\n",
    "        txt = ax.text(xtext, ytext, str(i), fontsize=24)\n",
    "        txt.set_path_effects([\n",
    "            PathEffects.Stroke(linewidth=5, foreground=\"w\"),\n",
    "            PathEffects.Normal()])\n",
    "        txts.append(txt)\n",
    "\n",
    "    return f, ax, sc, txts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(<matplotlib.figure.Figure at 0x7f16634d35d0>,\n",
       " <matplotlib.axes._subplots.AxesSubplot at 0x7f16634d34d0>,\n",
       " <matplotlib.collections.PathCollection at 0x7f1660de9c10>,\n",
       " [<matplotlib.text.Text at 0x7f169fa58c90>,\n",
       "  <matplotlib.text.Text at 0x7f1660dfdad0>,\n",
       "  <matplotlib.text.Text at 0x7f1660e0b0d0>,\n",
       "  <matplotlib.text.Text at 0x7f1660e0b690>,\n",
       "  <matplotlib.text.Text at 0x7f1660e0bc50>,\n",
       "  <matplotlib.text.Text at 0x7f1660e17250>,\n",
       "  <matplotlib.text.Text at 0x7f1660e17810>,\n",
       "  <matplotlib.text.Text at 0x7f1660eadf50>,\n",
       "  <matplotlib.text.Text at 0x7f1660ea9590>,\n",
       "  <matplotlib.text.Text at 0x7f1660f79350>])"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfMAAAHXCAYAAACoMrNrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nmAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f16634d35d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "scatter(digits_proj, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def _joint_probabilities_constant_sigma(D, sigma):\n",
    "    P = np.exp(-D**2/2 * sigma**2)\n",
    "    P /= np.sum(P, axis=1)\n",
    "    return P\n",
    "\n",
    "# Pairwise distances between all data points.\n",
    "D = pairwise_distances(X, squared=True, n_jobs =-1)\n",
    "# Similarity with constant sigma.\n",
    "P_constant = _joint_probabilities_constant_sigma(D, .002)\n",
    "# Similarity with variable sigma.\n",
    "P_binary = _joint_probabilities(D, 30., False)\n",
    "# The output of this function needs to be reshaped to a square matrix.\n",
    "P_binary_s = squareform(P_binary)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = np.array([[0, 2, 3],\n",
    "       [2, 0, 6],\n",
    "       [3, 6, 0]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0, 4, 5, 0, 4, 5],\n",
       "       [4, 0, 8, 4, 0, 8],\n",
       "       [5, 8, 0, 5, 8, 0]])"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spatial.distance.cdist(x, np.vstack([x, x])).astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f165fa06dd0>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "ename": "MemoryError",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mMemoryError\u001b[0m                               Traceback (most recent call last)",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/IPython/core/formatters.pyc\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, obj)\u001b[0m\n\u001b[1;32m    305\u001b[0m                 \u001b[0;32mpass\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    306\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 307\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mprinter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    308\u001b[0m             \u001b[0;31m# Finally look for special method names\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    309\u001b[0m             \u001b[0mmethod\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_real_method\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprint_method\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/IPython/core/pylabtools.pyc\u001b[0m in \u001b[0;36m<lambda>\u001b[0;34m(fig)\u001b[0m\n\u001b[1;32m    238\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    239\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0;34m'png'\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mformats\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 240\u001b[0;31m         \u001b[0mpng_formatter\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfor_type\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mFigure\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mlambda\u001b[0m \u001b[0mfig\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mprint_figure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfig\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'png'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    241\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0;34m'retina'\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mformats\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0;34m'png2x'\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mformats\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    242\u001b[0m         \u001b[0mpng_formatter\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfor_type\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mFigure\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mlambda\u001b[0m \u001b[0mfig\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mretina_figure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfig\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/IPython/core/pylabtools.pyc\u001b[0m in \u001b[0;36mprint_figure\u001b[0;34m(fig, fmt, bbox_inches, **kwargs)\u001b[0m\n\u001b[1;32m    122\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    123\u001b[0m     \u001b[0mbytes_io\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mBytesIO\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 124\u001b[0;31m     \u001b[0mfig\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcanvas\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprint_figure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbytes_io\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkw\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    125\u001b[0m     \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mbytes_io\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgetvalue\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    126\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mfmt\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m'svg'\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/backend_bases.pyc\u001b[0m in \u001b[0;36mprint_figure\u001b[0;34m(self, filename, dpi, facecolor, edgecolor, orientation, format, **kwargs)\u001b[0m\n\u001b[1;32m   2198\u001b[0m                     \u001b[0morientation\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0morientation\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2199\u001b[0m                     \u001b[0mdryrun\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mTrue\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2200\u001b[0;31m                     **kwargs)\n\u001b[0m\u001b[1;32m   2201\u001b[0m                 \u001b[0mrenderer\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfigure\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_cachedRenderer\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2202\u001b[0m                 \u001b[0mbbox_inches\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfigure\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_tightbbox\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrenderer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/backends/backend_agg.pyc\u001b[0m in \u001b[0;36mprint_png\u001b[0;34m(self, filename_or_obj, *args, **kwargs)\u001b[0m\n\u001b[1;32m    543\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    544\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mprint_png\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfilename_or_obj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 545\u001b[0;31m         \u001b[0mFigureCanvasAgg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdraw\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    546\u001b[0m         \u001b[0mrenderer\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_renderer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    547\u001b[0m         \u001b[0moriginal_dpi\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdpi\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/backends/backend_agg.pyc\u001b[0m in \u001b[0;36mdraw\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    462\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    463\u001b[0m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 464\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfigure\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdraw\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrenderer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    465\u001b[0m         \u001b[0;32mfinally\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    466\u001b[0m             \u001b[0mRendererAgg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlock\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrelease\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/artist.pyc\u001b[0m in \u001b[0;36mdraw_wrapper\u001b[0;34m(artist, renderer, *args, **kwargs)\u001b[0m\n\u001b[1;32m     61\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mdraw_wrapper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0martist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     62\u001b[0m         \u001b[0mbefore\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0martist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 63\u001b[0;31m         \u001b[0mdraw\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0martist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     64\u001b[0m         \u001b[0mafter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0martist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     65\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/figure.pyc\u001b[0m in \u001b[0;36mdraw\u001b[0;34m(self, renderer)\u001b[0m\n\u001b[1;32m   1142\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1143\u001b[0m             mimage._draw_list_compositing_images(\n\u001b[0;32m-> 1144\u001b[0;31m                 renderer, self, dsu, self.suppressComposite)\n\u001b[0m\u001b[1;32m   1145\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1146\u001b[0m             \u001b[0mrenderer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclose_group\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'figure'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/image.pyc\u001b[0m in \u001b[0;36m_draw_list_compositing_images\u001b[0;34m(renderer, parent, dsu, suppress_composite)\u001b[0m\n\u001b[1;32m    137\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mnot_composite\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mhas_images\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    138\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mzorder\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ma\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mdsu\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 139\u001b[0;31m             \u001b[0ma\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdraw\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrenderer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    140\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    141\u001b[0m         \u001b[0;31m# Composite any adjacent images together\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/artist.pyc\u001b[0m in \u001b[0;36mdraw_wrapper\u001b[0;34m(artist, renderer, *args, **kwargs)\u001b[0m\n\u001b[1;32m     61\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mdraw_wrapper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0martist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     62\u001b[0m         \u001b[0mbefore\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0martist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 63\u001b[0;31m         \u001b[0mdraw\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0martist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     64\u001b[0m         \u001b[0mafter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0martist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     65\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/axes/_base.pyc\u001b[0m in \u001b[0;36mdraw\u001b[0;34m(self, renderer, inframe)\u001b[0m\n\u001b[1;32m   2424\u001b[0m             \u001b[0mrenderer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstop_rasterizing\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2425\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2426\u001b[0;31m         \u001b[0mmimage\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_draw_list_compositing_images\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrenderer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdsu\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2427\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2428\u001b[0m         \u001b[0mrenderer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclose_group\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'axes'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/image.pyc\u001b[0m in \u001b[0;36m_draw_list_compositing_images\u001b[0;34m(renderer, parent, dsu, suppress_composite)\u001b[0m\n\u001b[1;32m    137\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mnot_composite\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mhas_images\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    138\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mzorder\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ma\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mdsu\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 139\u001b[0;31m             \u001b[0ma\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdraw\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrenderer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    140\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    141\u001b[0m         \u001b[0;31m# Composite any adjacent images together\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/artist.pyc\u001b[0m in \u001b[0;36mdraw_wrapper\u001b[0;34m(artist, renderer, *args, **kwargs)\u001b[0m\n\u001b[1;32m     61\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mdraw_wrapper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0martist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     62\u001b[0m         \u001b[0mbefore\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0martist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 63\u001b[0;31m         \u001b[0mdraw\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0martist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     64\u001b[0m         \u001b[0mafter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0martist\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     65\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/image.pyc\u001b[0m in \u001b[0;36mdraw\u001b[0;34m(self, renderer, *args, **kwargs)\u001b[0m\n\u001b[1;32m    541\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    542\u001b[0m             im, l, b, trans = self.make_image(\n\u001b[0;32m--> 543\u001b[0;31m                 renderer, renderer.get_image_magnification())\n\u001b[0m\u001b[1;32m    544\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mim\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    545\u001b[0m                 \u001b[0mrenderer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdraw_image\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mgc\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ml\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mim\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/image.pyc\u001b[0m in \u001b[0;36mmake_image\u001b[0;34m(self, renderer, magnification, unsampled)\u001b[0m\n\u001b[1;32m    768\u001b[0m         return self._make_image(\n\u001b[1;32m    769\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_A\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbbox\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtransformed_bbox\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0maxes\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbbox\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmagnification\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 770\u001b[0;31m             unsampled=unsampled)\n\u001b[0m\u001b[1;32m    771\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    772\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_check_unsampled_image\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/root/anaconda2/lib/python2.7/site-packages/matplotlib/image.pyc\u001b[0m in \u001b[0;36m_make_image\u001b[0;34m(self, A, in_bbox, out_bbox, clip_bbox, magnification, unsampled, round_to_pixel_border)\u001b[0m\n\u001b[1;32m    376\u001b[0m                     \u001b[0;31m# out of masked arrays.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    377\u001b[0m                     \u001b[0;32mwith\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0merrstate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minvalid\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'ignore'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 378\u001b[0;31m                         \u001b[0mrgba\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwhere\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mA\u001b[0m \u001b[0;34m<\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnan\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# under data\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    379\u001b[0m                         \u001b[0mrgba\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m2\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwhere\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mA\u001b[0m \u001b[0;34m>\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnan\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# over data\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    380\u001b[0m                     \u001b[0;31m# Have to invert mask, Agg knows what alpha means\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mMemoryError\u001b[0m: "
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1660dafbd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 4))\n",
    "pal = sns.light_palette(\"blue\", as_cmap=True)\n",
    "\n",
    "plt.subplot(131)\n",
    "plt.imshow(D, interpolation='none', cmap=pal)\n",
    "plt.axis('off')\n",
    "plt.title(\"Distance matrix\", fontdict={'fontsize': 16})\n",
    "\n",
    "# plt.subplot(132)\n",
    "# plt.imshow(P_constant[::10, ::10], interpolation='none', cmap=pal)\n",
    "# plt.axis('off')\n",
    "# plt.title(\"$p_{j|i}$ (constant $\\sigma$)\", fontdict={'fontsize': 16})\n",
    "\n",
    "# plt.subplot(133)\n",
    "# plt.imshow(P_binary_s[::10, ::10], interpolation='none', cmap=pal)\n",
    "# plt.axis('off')\n",
    "# plt.title(\"$p_{j|i}$ (variable $\\sigma$)\", fontdict={'fontsize': 16})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spatial.distance.is_valid_y([1,2,3,4,1,3])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}